Northwest Arkansas Democrat-Gazette
Slick rails, speed cited in train wreck
Patchy communications among causes, report says Safety Board
The same slick rails stalling a sight-seeing train in 2014 prevented the train that collided with it from stopping, according to a National Transportation Safety Board report released Thursday.
A fall foliage sight-seeing train of the Arkansas & Missouri Railroad Co. lost traction near Brentwood in Crawford County on its way from Springdale to Van Buren on Oct. 16, 2014. A train sent to assist hit the tourist train head-on about 10:25 a.m. that morning. All 39 passengers and four members of crew from the two trains were taken to local hospitals and medical centers for injuries, though none were life-threatening. Estimated damage was $178,500, the report states.
“Event recorder data show that just before the collision the independent air brakes of the work train locomotive were fully applied when the locomotive was about 230 feet from the excursion train,” the report said. “However, the work train could not stop in 230 feet, and it struck the excursion train.” The assisting train was supposed to be going no faster than 20 mph under these conditions, the report says, but hit the stalled train at 24 mph despite constant braking for more than 75 yards.
According to the engineer of the sight-seeing train, who wasn’t named in the report: “The sight-seeing train stalled on track The National Transportation Safety Board is the federal agency responsible for investigating accidents in involving air travel and transportation, highways, shipping and boats, pipelines and railroads, and any accident involving transportation of hazardous materials. It traces its origins to 1926, when Congress set up an agency to investigate the causes of aircraft accidents. The boar was taken out of the federal Department of Transportation in 1974 to ensure its independence.
that was black and covered with crushed leafy organic material. This condition, known as black rail, reduces adhesion.”
Others factors contributing to the wreck included patchy communications, violation of safety regulations and a partially bypassed system for spreading sand on the rails automatically in low-traction situations on the assisting train, the report said.
“The report says what it says, and they are very good at what they do,” Ron Sparks, spokesman for A&M, said Thursday of the safety board. The railroad company implemented most of the recommendations in the report long ago, he said, and is installing an improved radio system to eliminate dead spots that previously had to be filled with cellphone use, he said.
A&M has added a second locomotive to all sight-seeing trains to increase power,
revised its qualifications for working on and operating locomotives, put in new record keeping to track crew hours, rewritten rules for dispatchers, and tightened up procedures for assisting stalled trains long before the transportation board report was issued, according to both Sparks and the report.
Radio communication was always a problem in the mountainous terrain A&M operates in, both the report and Sparks said. Recent advances in technology, such as systems to repeat and amplify signals in hard-to-reach spots, are part of the new system due to be fully in place by early September, Sparks said.
“The collision caused the lead trucks of both locomotives to derail and all four cars of the excursion train to separate from the locomotive,” the report said. “The passenger cars rolled about 67 feet north.” The travel of the cars was stopped by automatic brakes triggered by their separation from the locomotive. The only injury serious enough to be described in the report was of a railroad employee. The assisting train’s conductor jumped from the locomotive’s cab before impact and suffered back and ankle injuries requiring admittance to a hospital, the report says.
“One of the damaged passenger cars released about 40 gallons of diesel fuel from a generator fuel tank, but the fuel did not catch fire,” the report says. The full text of the report is available at the transportation board’s website at www.ntsb.gov/investigations/AccidentReports/ Reports/RAB1704.pdf
The sight-seeing train had one locomotive and four passenger cars. The train left Springdale at about 8 a.m., headed for Van Buren. The train lost traction once during the trip, traveled another twoand-quarter miles and then lost all traction because of slippery rails. Recorder data confirmed the train had no mechanical problems, the report says. After-accident testing of the system the train had for spraying sand on the rails when wheels start spinning also showed the system worked, the report says.
The last two cars of the train carried all but one of the passengers, who was in the second car with one of the conductors. No one was in the first car, and the locomotive crew were all in their cab at the time of the accident, the report says.
The two-man work train crew and their locomotive were near Winslow when they were called by the stalled train’s engineer by cellphone for assistance. The railroad’s on-duty dispatcher also communicated with crews of both trains by cellphone to get the stalled train moving.
A&M railroad rules require “all train movements on out-of-service maintenance-of-way track are made at restricted speed,” the report says. “The data from the work train locomotive event recorder show the locomotive reached a speed of 35 mph during the move from Winslow siding toward (mile marker) 368,” the location of the stalled train. This speed violates federal regulations.” The report also said those regulations set a maximum speed under these conditions of 20 mph and that the manner in which the train was dispatched didn’t conform to regulations.
Although all A&M trains are equipped with radios, “certain areas of the railroad had unreliable radio communications because of the mountainous terrain in those areas,” the report says. “Therefore, A&M employees in jobs that required reliable communications in the areas with unreliable radio coverage used company-provided cell phones instead of radios for railroad operations. These phones were to be used as needed and only for company business in areas with unreliable radio communications.”
“Using cell phone as the only form of communication directly affected the operation of the work train locomotive,” the report continues. “Cell phone records show that although the two locomotive engineers made calls to each other on the day of the accident, they did not call each other in the minutes before the impact. Had there been better radio communication, the work train and excursion train engineers and the dispatcher could have confirmed the location of the stopped excursion train. In addition, other A&M employees may have been able to hear the conversations and could have provided important information to both engineers, such as track conditions and visibility approaching the excursion train.”
Gaps in radio communications is a contributing factor in several recent accidents in other railroads in other states, the report notes.
Post-accident interviews confirmed the work train engineer believed the stalled train was stopped 2 miles farther north than it was. “Neither of the engineers used formal communication procedures during their two cell phone conversations,” such as repeating important parts of their messages to make sure the location information was clearly understood, the report said. Such procedures are routinely used by dispatchers who oversee train traffic, but the two engineers were communicating directly by cell phone without going through the dispatcher in the last conversations before the accident, the report said. The dispatcher, meanwhile, was overseeing the operations of six trains including these two.
The investigation found this wasn’t the first time the rescue train’s engineer had “recently been on duty far longer than the time limits allowed” under federal regulations. “Moreover, before he operated a train on the main track, he had been off duty for fewer hours than required by federal regulations.”
After-accident testing of the system that automatically drops sand on the rails to improve traction when wheels start to slip show that it was working on the sight-seeing train. However, inspection of the sanding system on the rescuing train found that “axle 1, the leading axle during the rescue move, was cut out, or bypassed,” the report says. “In this condition, the sanding system could not function. Once the system was restored and axle 1 was no longer cut out, the sanders functioned as designed.”
Sparks said he couldn’t comment on why that system was bypassed because he didn’t know the reason.